Changes in the frequency and intensity of concurrent extreme wind speed and precipitation days over China during 1981&ndash;2022

Wen, Kangmin; Shi, Jun

doi:10.5194/egusphere-2025-5893

Preprints

https://doi.org/10.5194/egusphere-2025-5893

Preprints

04 Dec 2025

| 04 Dec 2025

Changes in the frequency and intensity of concurrent extreme wind speed and precipitation days over China during 1981–2022

Kangmin Wen and Jun Shi

Abstract. Compound extreme wind speed and extreme precipitation days (CWPDs) can significantly impact production and living, socio-economies, and human health. However, most previous studies have used fixed percentile thresholds to determine wind speed and precipitation extremes. This study investigated the characteristics and dynamics of CWPDs during 1981–2022 in mainland China based on the time-varying daily thresholds by using daily maximum wind speed and precipitation observation data. The results indicated that CWPDs were most likely to occur in southeastern South China (SC), Hainan Province, the northwestern parts of the middle and lower reaches of Yangtze River (YR), some scattered areas in the central and eastern YR. Annual CWPDs decreased in mainland China during 1981–2010, and then showed an obvious upward trend after 2010. Different percentile thresholds had effects on the spatial pattern and change trend of CWPDs. Spatially, CWPDs decreased more in parts of eastern Southwest China (SWC), YR, southeastern North China (NC) and central Northeast China (NEC), but less in mid-northern NC and most of Northwest China (NWC). In most areas of mainland China, the CWPDs frequencies under the four thresholds all showed decreasing trends, as the threshold increased, the trends of decreased for CWPDs frequencies decreased. With the increase of the threshold, the range of CWPDs with stronger intensity further reduced. CWPDs intensities were more severe in eastern coastal areas of YR, mid-eastern SC, parts of eastern SWC, parts of central NEC, parts of northwestern NWC and mid-northern in Hainan Province. Annual CWPDs intensities changed obvious around early-to-mid 2010s in under four different thresholds. With the increase of the threshold, the weakening trends of annual CWPDs intensities further weakened and even slightly strengthened during 1981–2022. The CWPDs intensities under the four thresholds all showed decreasing trends in most areas of mainland China except for parts of central SC, a few scattered areas of YR, several scattered areas of NC and NEC, a few scattered areas of NWC, individual areas of eastern SWC. As the thresholds increased, the trends of weakened for CWPDs intensities decreased and the scopes with a slight strengthening trend expanded. The changes of extreme wind speed days were consistent with those of CWPDs during 1981–2010 and 2011–2022, but the changes of extreme precipitation days and CWPDs were not corresponding. Due to the increase of extreme wind speed days and the accelerated increase of extreme precipitation days after 2010, the CWPDs changed from decrease before 2010 to increase after then. We conclude that the annual cumulative value obtained through time-varying thresholds and the latest daily observations can yield new insights into compound extremes.

Received: 27 Nov 2025 – Discussion started: 04 Dec 2025

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Preprint (PDF, 2637 KB)

Supplement (649 KB)

Download & links

Kangmin Wen and Jun Shi

Status: final response (author comments only)

RC1:
'Comment on egusphere-2025-5893', Anonymous Referee #1, 15 Dec 2025
Summary: The paper presents a detailed station-based analysis of compound extreme wind speed and extreme precipitation days (CWPDs) across mainland China using time-varying percentile thresholds. While the dataset is valuable and the results are detailed, the manuscript is overly descriptive. The motivation for time-varying thresholds and for selecting four specific thresholds is not clearly explained. Results emphasise detailed regional trends limiting broader national-scale trends. Overall, the study has potential value, but better integration of results across metrics, explicit rationale for threshold choices, and a focused discussion of post-2010 changes would improve readability and scientific impact.
Abstract
The abstract mentions time-varying daily thresholds but does not explain why fixed thresholds are insufficient or how time-varying thresholds improve the analysis. A brief reasoning would be helpful.

While threshold choice is emphasised, the abstract does not explain the four percentile thresholds, their physical or statistical meaning, or whether they are evaluated relative to fixed thresholds. Again, a brief justification would be useful for the readers.

The abstract is overly descriptive, with extensive regional detail. Consider synthesising the results into a few clear, high-level findings.

Trend descriptions are often qualitative (e.g., “obvious,” “slight”) and lack statistical context. Including basic metrics (e.g., trend magnitude, significance) would improve clarity.

The relationship between extreme wind, extreme precipitation, and CWPDs is not clearly articulated in the abstract and could be stated more explicitly.

The repeated emphasis on a shift in the early-to-mid 2010s lacks interpretation. A brief indication of whether this reflects physical drivers, data characteristics, or methodological choices would be helpful.

The reference to “annual cumulative value obtained through time-varying thresholds” introduces a concept not explained earlier in the abstract and may confuse readers. Consider clarifying or omitting this statement.

Introduction
The motivation in the opening para (lines 71–77) is repetitive and remains very general, relying on broad statements about global warming and compound extremes. The specific scientific problem addressed by this study is not clearly articulated until much later, and the motivation is not well linked to the metrics, region (China), or time scales analysed.

Previous studies are presented largely as a sequential list, with limited synthesis. It would benefit from clearer differentiation between studies focusing on frequency vs intensity, observational vs modelling approaches, and regional vs global perspectives, to better define what is known and where gaps remain.

The statement (lines 117-120) that gridded and reanalysis data are “less accurate than ground-based observations” is overly simplistic. Reanalysis products have known strengths and limitations, and a more balanced discussion would improve the credibility of this argument.

The claim that earlier studies focused mainly on single seasons is not fully convincing as currently presented. The Introduction does not clearly explain why all-season analysis is particularly important for compound wind-precipitation extremes.

While the limitations of using stable relative percentile thresholds are noted, the scientific rationale for adopting time-varying thresholds is not clearly explained.

The key research gap: limited use of time-varying thresholds with station-based observations in China is introduced very late in the Introduction (lines 115–124). This gap should be stated explicitly and earlier in the motivation.

The stated objective is technically clear but remains largely descriptive. Framing the study around specific research questions or testable hypotheses would strengthen the scientific focus.

Data and Methods
The choice of percentile thresholds (85^th, 90^th, 95^th, 98^th) is mentioned, but a justification of using these would strengthen reproducibility and interpretation.

The criterion allowing up to 20% missing data per station is not justified. Are stations with missing data skipped, or is some interpolation applied?

When aggregating daily CWPDI across all stations and selecting the top 10 days, why was the number 10 chosen? Does this capture meaningful spatial variability, or could a dense network of eastern stations dominate it?

What is the rationale behind focusing on W85∩P85 over higher percentiles? Could higher-threshold extremes show different dependency patterns?

Results
Section 3.1:
Four percentile thresholds are analysed equally, but the Results do not explain which threshold best represents CWPD behaviour, whether higher thresholds provide additional insight, or whether they mainly reduce sample size. An explicit empirical rationale would improve clarity.

It would be helpful to provide confidence intervals or summary statistics of spatial heterogeneity of the trends are reported as precise values (e.g., days per decade) (lines 257-259, 315-326).

Overall, it would be helpful to reduce the length of this section and highlight the key findings.

Section 3.2
The section provides extensive spatial and temporal description, but there is no concise summary of the dominant CWPD intensity behaviour. It is not explicitly stated whether intensity is overall increasing/decreasing, becoming more extreme at higher thresholds, or increasingly concentrated in specific regions.

The text does not clearly synthesise how intensity behaviour changes across thresholds (e.g., how many stations shift from decreasing to increasing trends, or whether high-threshold intensity changes are fundamentally different).

CWPD frequency and intensity are both central metrics, yet no explicit comparison is made. It would be good to know whether intensity and frequency trends are consistent, divergent, or regionally contrasting.

Section 3.3
The section title refers to “dependence,” but only co-variability and trend correspondence are shown. No explicit statistical linkage (e.g., correlation, sensitivity) between CWPDs and individual extremes is quantified. Please clarify early in the section whether the aim is to assess proportional contribution, trend coherence, or statistical dependence.

The Results imply dependence through ratios of CWPDs to individual extremes, but it is not explicitly stated why ratios are the appropriate metric?

Ratios are reported as % per decade, but mean ratio values and baseline magnitudes are not provided. Without this context, it is difficult to assess physical relevance or interpret changes.

Discussion
The Discussion section effectively emphasises time-varying thresholds and the 1981–2010 decrease / post-2010 increase in CWPDs, but it jumps between thresholds, spatial patterns, and synoptic systems. It would be helpful to restructure into: (a) key findings, (b) interpretation linked to results, (c) comparison with literature, and (d) implications and limitations.

The text introduces atmospheric rivers, tropical cyclones, Rossby waves, ENSO, PDO, AMO, NAO, urbanisation, aerosols, vegetation, and monsoon dynamics as explanations. Since these are not analysed in the study, suggest rephrasing as “plausible factors based on literature” or “may contribute to observed CWPD patterns”, and clearly distinguish them from findings supported by the current analysis. (Lines 607-623, 632-657, 658-718)
At present, much of this section reiterates established mechanisms with limited reference back to specific figures and/or quantitative findings in this paper.

Lines 736-739 about climate resilience and adaptation go beyond the analyses presented. Consider presenting these points as broader implications motivated by the findings, rather than as conclusions directly supported by the study.

The final paragraph clearly acknowledges the main methodological limitations of the study.

Conclusion
The section is largely consistent with the Results. Still, they are too long, and they read more like a condensed Results section (e.g., regional patterns and trends could be substantially reduced in favour of higher-level insights).

What is the single most important finding of this study that readers should remember?

What does this study add beyond previous CWPD research, particularly those using fixed thresholds?

Why do time-varying thresholds matter scientifically, not just methodologically?

Why is the post-2010 shift in CWPD frequency and intensity important in a broader climate context?

How do these findings improve understanding of compound wind–precipitation extremes under climate change?
Citation: https://doi.org/10.5194/egusphere-2025-5893-RC1
- AC1: 'Reply on RC1', K.M. Wen, 06 Mar 2026
  
  Summary: The paper presents a detailed station-based analysis of compound extreme wind speed and extreme precipitation days (CWPDs) across mainland China using time-varying percentile thresholds. While the dataset is valuable and the results are detailed, the manuscript is overly descriptive. The motivation for time-varying thresholds and for selecting four specific thresholds is not clearly explained. Results emphasise detailed regional trends limiting broader national-scale trends. Overall, the study has potential value, but better integration of results across metrics, explicit rationale for threshold choices, and a focused discussion of post-2010 changes would improve readability and scientific impact.
  Thank you very much for the comments. We have made a new revision of the manuscript, referring to the comments and suggestion.
  
  Abstract
  
  (1) The abstract mentions time-varying daily thresholds but does not explain why fixed thresholds are insufficient or how time-varying thresholds improve the analysis. A brief reasoning would be helpful.
  Thank you for the comment. In the abstract section of this revised manuscript, we have changed “time-varying thresholds” to “thresholds that vary with dates and stations”. In addition, in the introduction of the revised manuscript, we have explained why fixed thresholds are insufficient or how time-varying thresholds improve the analysis: “Most of the previous studies have calculated the compound extremes according to the stable relative percentile thresholds, and the analysis of the compound wind speed and precipitation extremes base on the thresholds that vary with dates and stations is lacking. Since there are significant differences in wind speed and precipitation among different seasons in China, it is obviously inappropriate to choose a fixed threshold for extreme wind speed and extreme precipitation to analyze the climate changes caused by extreme wind speed and extreme precipitation. Similarly, because the wind speed and precipitation vary greatly in different parts of China, it is obviously inappropriate to use the same fixed extreme wind speed threshold and extreme precipitation threshold for all stations across the country to analyze climate change caused by extreme wind speed and extreme precipitation. Therefore, we use the thresholds that vary with dates and stations to define compound wind speed and precipitation extremes”.
  
  (2) While threshold choice is emphasised, the abstract does not explain the four percentile thresholds, their physical or statistical meaning, or whether they are evaluated relative to fixed thresholds. Again, a brief justification would be useful for the readers.
  We accept your suggestion. In the data and methods section of this revised manuscript, we have referred to more references to illustrate their suitability in capturing compound extremes and the percentiles chosen are designed to serve as a trade-off between extremeness and sample size.
  
  (3) The abstract is overly descriptive, with extensive regional detail. Consider synthesising the results into a few clear, high-level findings.
  Thank you for the suggestion. We have resummarized the abstract as: “Compound extreme wind speed and extreme precipitation days (CWPDs) can significantly impact production and living, socio-economies, and human health. However, most previous studies have used fixed percentile thresholds to determine wind speed and precipitation extremes. This study investigated the characteristics and dynamics of CWPDs during 1981–2022 in mainland China based on the thresholds that vary with dates and stations by using daily maximum wind speed and precipitation observation data. The results indicated that different percentile thresholds had effects on the spatial pattern and change trend of CWPDs. The CWPDs frequencies were most likely to occur in southeastern China and some areas of the middle and lower reaches of Yangtze River (YR), the CWPDs intensities were more severe in southeastern China, and some areas include eastern SWC and northwestern NWC. In most areas of mainland China, the CWPDs frequencies under the four thresholds all showed decreasing trends, the CWPDs intensities under the four thresholds also showed decreasing trends except for some parts of southeastern China, as the thresholds increased, the CWPDs frequencies distribution patterns are mostly consistent, but the change trends decreased at the same location, the weakened trends and areas of the CWPDs intensities decreased and the areas with a slight strengthening trend expanded. For mainland China as a whole, with the heightening of the threshold, the decreasing trend of annual CWPDs frequencies had slowed down, and the weakening trends of annual CWPDs intensities further weakened and even slightly strengthened during 1981-2022. In the past 42 years, the CWPDs frequencies under the four thresholds of W85∩P85, W90∩P90, W95∩P95 and W98∩P98 decreased significantly at a rate of 0.54, 0.18, 0.10 and 0.02 days per decade, respectively. The CWPDs intensities under the thresholds of W85∩P85 and W90∩P90 decreased significantly at a rate of 3.59 and 1.47 per decade, but under the threshold of W95∩P95 decreased non-significantly at a rate of 0.07 per decade and under the threshold of W98∩P98 increased non-significantly at a rate of 0.01 per decade. Both the frequencies and the intensities of CWPDs showed a downward trend in 1981–2010 followed by an upward trend in 2011–2022 under the four thresholds. The changes of extreme wind speed days were consistent with those of CWPDs during 1981–2010 and 2011–2022, but the changes of extreme precipitation days and CWPDs were not corresponding. Compared to extreme precipitation days, the CWPDs is dominated by extreme wind speed days. CWPDs and extreme wind speed days are decreased simultaneously during 1981 to 2010. Due to the increase of extreme wind speed days and the accelerated increase of extreme precipitation days after 2010, the CWPDs changed from decrease before 2010 to increase after then. We conclude that by applying thresholds that vary with dates and stations and the latest daily observations can yield new insights into compound extremes”.
  
  (4) Trend descriptions are often qualitative (e.g., “obvious,” “slight”) and lack statistical context. Including basic metrics (e.g., trend magnitude, significance) would improve clarity.
  Thank you for the suggestion. In the abstract section of the revised manuscript, we have added quantitative trends magnitude and significance.
  
  (5) The relationship between extreme wind, extreme precipitation, and CWPDs is not clearly articulated in the abstract and could be stated more explicitly.
  Thank you for the suggestion. In the abstract section of the revised manuscript, we have reinterpreted the relationship between extreme wind, extreme precipitation and CWPDs: “The changes of extreme wind speed days were consistent with those of CWPDs during 1981–2010 and 2011–2022, but the changes of extreme precipitation days and CWPDs were not corresponding. Compared to extreme precipitation days, the CWPDs was dominated by extreme wind speed days. CWPDs and extreme wind speed days were decreased simultaneously during 1981 to 2010, although precipitation extremes were showing a decreasing trend during this period. Due to the increase of extreme wind speed days and the accelerated increase of extreme precipitation days after 2010, the CWPDs changed from decrease before 2010 to increase after then”.
  
  (6) The repeated emphasis on a shift in the early-to-mid 2010s lacks interpretation. A brief indication of whether this reflects physical drivers, data characteristics, or methodological choices would be helpful.
  Thank you for the suggestion. In the abstract section of the revised manuscript, we have described this data characteristic as: “Compared to extreme precipitation days, the CWPDs was dominated by extreme wind speed days. CWPDs and wind speed extremes were decreased simultaneously during 1981 to 2010, although precipitation extremes were showing a decreasing trend during this period. Due to the increase of extreme wind speed days and the accelerated increase of extreme precipitation days after 2010, the CWPDs changed from decrease before 2010 to increase after then”.
  
  (7) The reference to “annual cumulative value obtained through time-varying thresholds” introduces a concept not explained earlier in the abstract and may confuse readers. Consider clarifying or omitting this statement.
  Thank you for the suggestion. In the abstract section of the revised manuscript, we have changed the“time-varying thresholds” to “thresholds that vary with dates and stations”.
  
  Introduction
  
  (8) The motivation in the opening para (lines 71–77) is repetitive and remains very general, relying on broad statements about global warming and compound extremes. The specific scientific problem addressed by this study is not clearly articulated until much later, and the motivation is not well linked to the metrics, region (China), or time scales analysed.
  Thank you for the suggestion. In the introduction section of the revised manuscript, we have deleted the repetitive motivation, rewrited the motivation and moved the specific scientific problem to the front: “China is one of the countries with the highest frequency of occurrence of compound extreme events (Liu et al., 2023). The frequency, duration and intensity of extreme weather and climate events in most parts of China have further increased after the 1970s (Ren et al., 2014; Sun et al., 2019; Wu et al., 2016). The existing studies on compound extreme events in China mostly focus on compound warm-dry events (Feng et al., 2020; Kong et al., 2020; Haqiqi et al., 2021; Li et al., 2019; Zhang et al., 2021), while there are relatively few studies on compound wind-precipitation events”.
  
  (9) Previous studies are presented largely as a sequential list, with limited synthesis. It would benefit from clearer differentiation between studies focusing on frequency vs intensity, observational vs modelling approaches, and regional vs global perspectives, to better define what is known and where gaps remain.
  Thank you for the suggestion. In the introduction section of the revised manuscript, we have resummarized the literatures focusing on frequency vs intensity, observational vs modelling approaches, and regional vs global perspectives.
  
  (10) The statement (lines 117-120) that gridded and reanalysis data are “less accurate than ground-based observations” is overly simplistic. Reanalysis products have known strengths and limitations, and a more balanced discussion would improve the credibility of this argument.
  Thank you for the suggestion. In the introduction section of the revised manuscript, we have redescribed the flaws of reanalysis products: In some regions with complex terrain (e.g., mountainous areas), reanalysis datas are not suitable as a benchmark for studying compound wind speed and precipitation extremes (Zscheischler et al., 2020b; Zhang et al., 2021). Observational records is crucial for accurately, comprehensively investigating the authentic characteristics of compound wind speed and precipitation extremes”.
  
  (11) The claim that earlier studies focused mainly on single seasons is not fully convincing as currently presented. The Introduction does not clearly explain why all-season analysis is particularly important for compound wind-precipitation extremes.
  Thank you for the suggestion. In the introduction section of the revised manuscript, we have reinterpreted all-season analysis is particularly important for compound wind-precipitation extremes: “Compound wind speed and precipitation extremes are often drived by tropical cyclones (e.g., typhoons) (Chen et al., 2019b; Raymond et al., 2020) in summer and autumn, or strong convective weather systems (Lukens et al., 2018; Yu and Zhong, 2019) in spring and summer, or the landfall of atmospheric rivers (Waliser and Guan, 2017). Besides, special terrain and other low-pressure systems (Martius et al., 2016; Zscheischler et al., 2020a) can trigger compound wind speed and precipitation extremes in all seasons. Therefore, all-season analysis is particularly important for compound wind speed and precipitation extremes”.
  
  (12) While the limitations of using stable relative percentile thresholds are noted, the scientific rationale for adopting time-varying thresholds is not clearly explained.
  Thanks for the suggestion. In the revised manuscript, we have added the description of the motivation for using thresholds that vary with dates and stations in the introduction section “Since there are significant differences in wind speed and precipitation among different seasons in China, it is obviously inappropriate to choose a fixed threshold for extreme wind speed and extreme precipitation to analyze the climate changes caused by extreme wind speed and extreme precipitation. Similarly, because the wind speed and precipitation vary greatly in different parts of China, it is obviously inappropriate to use the same fixed extreme wind speed threshold and extreme precipitation threshold for all stations across the country to analyze climate change caused by extreme wind speed and extreme precipitation. Therefore, we use the thresholds that vary with dates and stations to define compound wind speed and precipitation extremes”.
  
  (13) The key research gap: limited use of time-varying thresholds with station-based observations in China is introduced very late in the Introduction (lines 115–124). This gap should be stated explicitly and earlier in the motivation.
  Thanks for the suggestion. In the introduction section of the revised manuscript, we have stated the gap explicitly and earlier in the motivation.
  
  (14) The stated objective is technically clear but remains largely descriptive. Framing the study around specific research questions or testable hypotheses would strengthen the scientific focus.
  Thanks for the suggestion. In the introduction section of the revised manuscript, we have redescribed the objective: “The objective of the present study therefore is to (a) investigate the spatial and temporal characteristics, including frequency and intensity of compound extreme wind speed and extreme precipitation days based on the daily observational records from ground stations, (b) quantitatively estimate the effects of wind speed or precipitation extremes on compound wind speed and precipitation extremes or the dependencies of compound extremes on individual hot or dry extremes, and (c) analyze the influence characteristics of research scales and prescribed thresholds on compound extremes based on the thresholds that vary with dates and stations. The results of this study may be useful in providing support for the government’s decision-making on agricultural management, energy distribution, regional stability maintenance, transportation protection, ecological environmental protection, and reducing the threat of compound wind speed and precipitation extremes to human lives and properties”.
  
  Data and Methods
  
  (15) The choice of percentile thresholds (85th, 90th, 95th, 98th) is mentioned, but a justification of using these would strengthen reproducibility and interpretation.
  Thanks for the suggestion. In this revised manuscript, we have referred to more references to illustrate their suitability in capturing compound extremes and the percentiles chosen are designed to serve as a trade-off between extremeness and sample size.
  
  (16) The criterion allowing up to 20% missing data per station is not justified. Are stations with missing data skipped, or is some interpolation applied?
  Thanks for the suggestion. In the data and methods section of this revised manuscript, we have redescribed the data processing procedure: “The relevant data should coexist at a given station in the same time series to facilitate the study of compound wind speed and precipitation extreme characteristics. The maximum wind speed records at many stations had missing values before 1981, so to ensure the consistency and integrity of the daily sequences, if the missing rate of daily records at a station exceeded 20% of the total daily records from 1981 to 2022, the meteorological station would be removed from the research dataset. Ultimately, 1686 meteorological stations were selected for analysis (Fig. 1). Among the 1686 stations, there were 1292 stations with the missing rate of daily maximum wind speed records of less than 3%, and 1654 stations with that of daily precipitation records of less than 3% during 1981–2022. The missing values of wind speed and precipitation in any station were replaced by the records from the nearest station on the same day”.
  
  (17) When aggregating daily CWPDI across all stations and selecting the top 10 days, why was the number 10 chosen? Does this capture meaningful spatial variability, or could a dense network of eastern stations dominate it?
  Thanks for the suggestion. For the the top 10 daily CWPDI didn’t analyze in present study, in the data and methods section of this revised manuscript, we have deleted the sentence: “We summed the CWPDI values of all stations in China on each day (1981–2022) to determine the daily cumulative CWPDI values in descending order, ultimately isolating the top 10 concurrent extremes for further analysis”.
  
  (18) What is the rationale behind focusing on W85∩P85 over higher percentiles? Could higher-threshold extremes show different dependency patterns?
  Thanks for the suggestion. At the end of the data and methods section of the revised manuscript, we have restated the reasons for choosing the threshold of W85∩P85: “For a nationwide analysis, a higher threshold will result in a smaller sample size in inland areas of the country, and the threshold of W85∩P85 can ensure an adequate sample size to reflect the differences in CWPD variations across different regions. Therefore, the threshold of W85∩P85 was used for further explore seasonal characteristics of CWPDs, and the relationship between extreme wind speed, extreme precipitation and CWPDs”.
  
  Results
  
  Section 3.1:
  
  (19) Four percentile thresholds are analysed equally, but the Results do not explain which threshold best represents CWPD behaviour, whether higher thresholds provide additional insight, or whether they mainly reduce sample size. An explicit empirical rationale would improve clarity.
  Thank you for the comment. In the section 3.1 of this revised manuscript, we have added the explicit explanation: “In contrastively, the threshold of W85∩P85 could have more samples, and identify frequent CPWDs for a wider regions nationwide, as the thresholds increasesd, the sample size reduced, the identified frequent CPWDs only concentrated in the southeastern coastal areas of the country”.
  
  (20) It would be helpful to provide confidence intervals or summary statistics of spatial heterogeneity of the trends are reported as precise values (e.g., days per decade) (lines 257-259, 315-326). Overall, it would be helpful to reduce the length of this section and highlight the key findings.
  Thanks for the suggestion. In the section 3.1 of this revised manuscript, we have provided confidence intervals of the precise trends, and reduced the length of this section and highlight the key findings.
  
  Section 3.2
  
  (21) The section provides extensive spatial and temporal description, but there is no concise summary of the dominant CWPD intensity behaviour. It is not explicitly stated whether intensity is overall increasing/decreasing, becoming more extreme at higher thresholds, or increasingly concentrated in specific regions.
  Thanks for the suggestion. In the section 3.2 of this revised manuscript, we have summarized the dominant behaviour of CWPD intensity: “In mainland China, CWPDs were less serious in most NEC, NWC, SWC and NC, but severe in eastern coastal areas of YR, mid-eastern SC, parts of eastern SWC, parts of central NEC, parts of northwestern NWC and mid-northern in Hainan Province. Different percentile thresholds had certain effects on the spatial pattern of the intensity of CWPDs, with the increase of the threshold, the area of CWPDs with stronger intensity further reduced, and stronger CPWDs concentrated in the southeastern coastal areas of the country”.
  
  (22) The text does not clearly synthesise how intensity behaviour changes across thresholds (e.g., how many stations shift from decreasing to increasing trends, or whether high-threshold intensity changes are fundamentally different).
  Thanks for the suggestion. In the section 3.2 of this revised manuscript, we have synthesised how intensity behaviour changes across thresholds.
  
  (23) CWPD frequency and intensity are both central metrics, yet no explicit comparison is made. It would be good to know whether intensity and frequency trends are consistent, divergent, or regionally contrasting.
  Thanks for the suggestion. In the section 3.2 of this revised manuscript, we have compared the intensity with the frequency: “For mainland China as a whole, from 1981 to 2022, the frequencies of the four thresholds all showed a downward trend. As the threshold increases, the downward trend becomes smaller. The intensities of the W85∩P85 and W90∩P90 thresholds showed a significant downward trend, while the intensities of the W95∩P95 and W98∩P98 thresholds did not change significantly. When looking at the time periods separately, both frequency and intensity showed a weakening trend from 1981 to 2010, and an increasing trend from 2011 to 2022. In terms of spatial distribution, the trends of frequency and intensity are generally consistent, and both are the area showing a decreasing trend larger than the area showing an increasing trend. As the threshold increases, the spatial trend distribution of CWPDs frequency does not change significantly at W85∩P85, W90∩P90, and W95∩P95 thresholds. When the threshold was raised to W98∩P98 percentiles, the area of frequency showing an increasing trend significantly expands. The spatial trend distribution of CWPDs intensity does not change much at W85∩P85 and W90∩P90 thresholds, when the thresholds were raised to W95∩P95 and W98∩P98 percentiles, the area of intensity showing an enhanced trend significantly expands. In terms of the W98∩P98 threshold, the stations with an increasing trend in intensity and those with a decreasing trend are almost the same”.
  
  Section 3.3
  
  (24) The section title refers to “dependence,” but only co-variability and trend correspondence are shown. No explicit statistical linkage (e.g., correlation, sensitivity) between CWPDs and individual extremes is quantified. Please clarify early in the section whether the aim is to assess proportional contribution, trend coherence, or statistical dependence.
  Thanks for the suggestion. In the beginning of the section 3.3 of this revised manuscript, we have added the explanation: “To explored the relationship between extreme wind speed, extreme precipitation and CWPDs, the proportional contribution and trend coherence of concurrent days on individual ones were further analyzed”. In addition, we have changed the title of section 3.3 to “Proportional contribution and trend coherence of concurrent days on individual ones”.
  
  (25) The Results imply dependence through ratios of CWPDs to individual extremes, but it is not explicitly stated why ratios are the appropriate metric?
  Thanks for the suggestion. In the section 3.3 of this revised manuscript, we have added the explanation: “We divided the CWPDs by the corresponding precipitation and wind speed extremes (i.e., their ratio) to evaluate the linkage between individual and concurrent extremes reference the method of Zhang et al., 2021.”.
  
  Reference:
  Zhang, Y., Sun, X., Chen, C.: Characteristics of concurrent precipitation and wind speed extremes in China, Weather and Climate Extremes, 32, 100322, 2021.
  
  (26) Ratios are reported as % per decade, but mean ratio values and baseline magnitudes are not provided. Without this context, it is difficult to assess physical relevance or interpret changes.
  Thanks for the suggestion. In the section 3.3 of this revised manuscript, we have added mean ratio values and baseline magnitudes: “For mainland China as a whole, the country average ratios of CWPDs to extreme wind speed days during 1981-2022, 1981-2010 and 2011-2022 were 10.4%, 9.5% and 12.6%, respectively. For extreme precipitation days, the country average ratios during 1981-2022, 1981-2010 and 2011-2022 were 26.1%, 28.0% and 21.4%, respectively (Fig. 10b)”.
  
  Discussion
  
  (27) The Discussion section effectively emphasises time-varying thresholds and the 1981–2010 decrease / post-2010 increase in CWPDs, but it jumps between thresholds, spatial patterns, and synoptic systems. It would be helpful to restructure into: (a) key findings, (b) interpretation linked to results, (c) comparison with literature, and (d) implications and limitations.
  Thanks for the suggestion. In the discussion section of this revised manuscript, we have restructured the discussion into: (a) key findings, (b) interpretation linked to results and (c) implications and limitations. The content of comparison with literature was placed in the results section.
  
  (28) The text introduces atmospheric rivers, tropical cyclones, Rossby waves, ENSO, PDO, AMO, NAO, urbanisation, aerosols, vegetation, and monsoon dynamics as explanations. Since these are not analysed in the study, suggest rephrasing as “plausible factors based on literature” or “may contribute to observed CWPD patterns”, and clearly distinguish them from findings supported by the current analysis. (Lines 607-623, 632-657, 658-718)
  Thank you for the comment. In the discussion section of this revised manuscript, we have rephrasing the explanations as “may contribute to observed CWPD patterns” and “plausible factors based on literature”, and clearly distinguish them from findings of the current analysis.
  
  (29) At present, much of this section reiterates established mechanisms with limited reference back to specific figures and/or quantitative findings in this paper.
  Thanks for the suggestion. In the discussion section of this revised manuscript, we have added the description of specific figures and quantitative findings of the current analysis.
  
  (30) Lines 736-739 about climate resilience and adaptation go beyond the analyses presented. Consider presenting these points as broader implications motivated by the findings, rather than as conclusions directly supported by the study.
  Thanks for the suggestion. We have presented the points as broader implications: “As compound extremes are more devastating and damaging than independent extreme event, in this study, we mainly concentrated on the wind speed and precipitation extremes that occur on the same day, namely, the concurrent extreme wind speed and extreme precipitation days, such incidents coupled with high exposure and vulnerability of the population or crops, causing huge losses in multiple aspects including water security, food security and human health (Seneviratne et al., 2012; Yaddanapudi et al., 2022). Our finding indicated that high-intensity composite wind speed and precipitation extremes have occurred frequently in the past few decades, especially in eastern coastal areas of YR, mid-eastern SC, parts of eastern SWC and mid-northern in Hainan Province. As the frequencies and intensities of composite extreme events increased in climate vulnerable areas after early-to-mid 2010s, indicating that this increasing trend will continue in the future, it is necessary to enhance climate adaptability or resilience in these regions, such as increasing water conservancy development in agricultural areas, change the agricultural production mode, and introduce drought-resistant and heat-tolerant crop varieties”.
  
  (31) The final paragraph clearly acknowledges the main methodological limitations of the study.
  Thanks for the suggestion. We have presented the main methodological limitations of the study: “It is worth noting that the missing values of wind speed and precipitation in any station were interpolated by the records from the nearest station on the same day in present study may result in the statistics of CWPDs not being completely accurate, besides, compound wind speed and precipitation extremes are not just phenomena that occur on the same day, and other forms should also be addressed. For instance, extreme wind speed or extreme precipitation occurs a day or two after or before CWPDs (Zscheischler et al., 2020), wind speed and precipitation extremes appear in a relatively short period of time in nearby regions (Raymond et al., 2020), and so on. Moreover, the physical process diagnosis, change attribution, risk quantification and adaptation measures of CWPDs are also not involved in this paper, which need to be explored in future studies”.
  
  Conclusion
  
  (32) The section is largely consistent with the Results. Still, they are too long, and they read more like a condensed Results section (e.g., regional patterns and trends could be substantially reduced in favour of higher-level insights).
  Thanks for the suggestion. We have substantially reduced the regional patterns and trends and provided more meaningful insights: “We investigated the observed CWPDs patterns in the past 42 years over mainland China by using the thresholds that vary with dates and stations, which could accurately provide information of CWPDs on different time scales and periods for different locations under the background of global warming. Our analysis identified the hotspot locations of CWPDs were in northwestern and eastern coastal areas of YR, mid-eastern and southeastern SC, Hainan Province, parts of eastern SWC, parts of central NEC, parts of northwestern NWC, which complements previous studies (Bevacqua et al., 2020; Couasnon et al., 2020; Martius et al., 2016). Both the frequencies and the intensities of CWPDs changed obvious around early-to-mid 2010s under four different thresholds, showed a downward trend in 1981–2010 followed by an upward trend in 2011–2022, indicating that high-intensity CWPDs will occur frequently in the future, especially in the hotspot locations, this is of great practical significance in guiding these regions to formulate corresponding climate adaptation and mitigation strategies. The CWPDs frequencies decreased during 1981-2010 were mainly caused by the decreased of extreme wind speed days, but the CWPDs frequencies increased during 2011-2022 were caused by the increased of both extreme wind speed days and extreme precipitation days. The changes of extreme wind speed days were consistent with those of CWPDs during 1981–2010 and 2011–2022, but the change of extreme precipitation days lacked correspondence with that of CWPDs in the two periods, especially in the previous period.
  Different percentile thresholds had effects on the spatial pattern and change trend of CWPDs. As the threshold increases, the CWPDs frequencies under the four thresholds all showed decreasing trends in most areas of mainland China during 1981 to 2022, the trends of decreased for CWPDs frequencies decreased obviously, the spatial trend distribution of CWPDs frequency does not change significantly at W85∩P85, W90∩P90, and W95∩P95 thresholds. When the threshold was raised to W98∩P98 percentiles, the area of frequency showing an increasing trend significantly expands. The spatial trend distribution of CWPDs intensity does not change much at W85∩P85 and W90∩P90 thresholds, when the thresholds were raised to W95∩P95 and W98∩P98 percentiles, the area of intensity showing an enhanced trend significantly expands. In terms of the W98∩P98 threshold, the stations with an increasing trend in intensity and those with a decreasing trend are almost the same.
  Our findings provide valuable information (e.g., hotspot locations, intensity analysis, and sensitivity analysis) related to the observed CWPDs patterns over mainland China. The findings will enhancement future researches on CWPDs dominated by different weather systems, such as tropical cyclones in southeastern coastal areas of mainland China (Rodgers et al., 1994) and stronger extratropical cyclones in Jianghuai region and Northeast region of mainland China (Raveh-Rubin & Wernli, 2015). The findings from our study can help reduce the risk from concurrent events for various stakeholders, such as coastal communities, utility industries, agriculture agency, emergency management agency, and the insurance sector”.
  
  (33) What is the single most important finding of this study that readers should remember?
  Thanks for the suggestion. In the conclusion section of this revised manuscript, we have rewrited the most important finding: “Different percentile thresholds had effects on the spatial pattern and change trend of CWPDs. As the threshold increases, the CWPDs frequencies under the four thresholds all showed decreasing trends in most areas of mainland China during 1981 to 2022, the trends of decreased for CWPDs frequencies decreased obviously, the spatial trend distribution of CWPDs frequency does not change significantly at W85∩P85, W90∩P90, and W95∩P95 thresholds. When the threshold was raised to W98∩P98 percentiles, the scope of frequency showing an increasing trend significantly expands. The spatial trend distribution of CWPDs intensity does not change much at W85∩P85 and W90∩P90 thresholds, when the thresholds were raised to W95∩P95 and W98∩P98 percentiles, the scope of intensity showing an enhanced trend significantly expands. In terms of the W98∩P98 threshold, the stations with an increasing trend in intensity and those with a decreasing trend are almost the same”.
  
  (34) What does this study add beyond previous CWPD research, particularly those using fixed thresholds?
  Thanks for the suggestion. At the beginning of the conclusion section of this revised manuscript, we have clearly stated the innovative aspects of this research compared to previous studies: “We investigated the observed CWPDs patterns in the past 42 years over mainland China by using the thresholds that vary with dates and stations, which could accurately provide information of CWPDs on different time scales and periods for different locations under the background of global warming”.
  
  (35) Why do time-varying thresholds matter scientifically, not just methodologically?
  Thanks for the suggestion. At the beginning of the conclusion section, we have explained the significance of using the thresholds that vary with dates and stations: “We investigated the observed CWPDs patterns in the past 42 years over mainland China by using the thresholds that vary with dates and stations, which could accurately provide information of CWPDs on different time scales and periods for different locations under the background of global warming”.
  
  (36) Why is the post-2010 shift in CWPD frequency and intensity important in a broader climate context?
  Thanks for the suggestion. At the end of the discussion section, we have explained the significance of the post-2010 shift in CWPD frequency and intensity: “As the frequencies and intensities of composite extreme events increased in climate vulnerable areas after early-to-mid 2010s, indicating that this increasing trend will continue in the future, it is necessary to enhance climate adaptability or resilience in these regions, such as increasing water conservancy development in agricultural areas, change the agricultural production mode, and introduce drought-resistant and heat-tolerant crop varieties”.
  
  (37) How do these findings improve understanding of compound wind–precipitation extremes under climate change?
  Thanks for the suggestion. At the end of the discussion section of this revised manuscript, we have stated the significance of the findings in present study to improve understanding of compound wind–precipitation extremes under climate change: “Our findings provide valuable information (e.g., hotspot locations, intensity analysis, and sensitivity analysis) related to the observed CWPDs patterns over mainland China. The findings will enhancement future researches on CWPDs dominated by different weather systems, such as tropical cyclones in southeastern coastal areas of mainland China (Rodgers et al., 1994) and stronger extratropical cyclones in Jianghuai region and Northeast region of mainland China (Raveh-Rubin & Wernli, 2015). The findings from our study can help reduce the risk from concurrent events for various stakeholders, such as coastal communities, utility industries, agriculture agency, emergency management agency, and the insurance sector”.
  
  Thanks again for your comments.
  
  Citation: https://doi.org/10.5194/egusphere-2025-5893-AC1
RC2:
'Comment on egusphere-2025-5893', Anonymous Referee #2, 02 Feb 2026

This manuscript investigates compound wind and precipitation events (CWPDs) using an extensive network of meteorological stations across China. Both the frequency and intensity of compound events are evaluated using different percentile-based threshold combinations. Overall, the manuscript is highly descriptive but lacks a more comprehensive interpretation of the results.
The connection between the methodology and the results should be strengthened. In particular, the selection of the time window and percentile thresholds applied to the time series requires clearer justification, as does the rationale for combining specific wind and precipitation thresholds. The manuscript would also benefit from additional analyses addressing the seasonality of CWPDs, as well as spatial correlations and/or trends.
Comments
1 Introduction.
There is a strong emphasis on compound wind–precipitation events in coastal areas throughout the introduction; however, the stated objective clearly indicates that the focus is mainland China. This inconsistency should be addressed.
119: “so the results are less accurate than ground-based observations” - This statement is not necessarily correct. Reanalysis and model data are typically calibrated and validated against ground-based observations.
2 Data
This section should be titled “Data and Methods” rather than only “Data.”
In several cases, the methods lack sufficient explanation or justification for key choices (e.g., threshold selection). Additionally, some methods are mentioned without even a brief explanation, while others are described in detail. The linkage between the methods and the results could be improved.
2.1 Data
144: “The missing records of wind speed and precipitation in any station were retained.” - How were these missing records handled in the analysis?
Overall, a clearer overview of the dataset is needed, including the number of days with wind events, precipitation events, and CWPDs.
2.2 Definition of compound extremes
166: The justification of the 15-day time window is missing
168-169: This sentence is not clear
173: The choice of percentile thresholds (85th, 90th, 95th, 98th) requires justification
179: The rationale for the selected threshold combinations is missing.
196: “ultimately isolating the top 10 concurrent extremes for further analysis.” – Is this actually done? The results appear to include all events. If not, a clearer explanation is needed.
2.3 Regional division and linear trend
199: It would be helpful to indicate how many stations are included in each region (e.g., in Figure 2).
209: Briefly describe the Sen–Theil method.
219: Why was the W85∩P85 threshold combination selected?
220: Please explain how the frequency ratios are defined and calculated.
3 Results
Hainan Province is mentioned multiple times without being introduced. It should be clearly identified in Figure 1.
Information regarding threshold intensity is missing. For example, which actual wind speeds and precipitation amounts correspond to the different percentiles? Are there clear spatial patterns in these intensities?
Figures and tables are described extensively, but there is limited discussion of the physical mechanisms or spatial patterns underlying the results.
An analysis by region may be more informative.
3.1 Characteristics of CWPDs frequency
Figure 2: Use consistent color scales for all four threshold combinations. The region names are difficult to read; consider moving them to the foreground or outside the map.
It is expected that similar spatial patterns will emerge across different thresholds, as they are percentile-based. Much of the text reiterates what is already visible in the figures and becomes repetitive when discussing multiple threshold combinations.
Instead, the discussion should focus more on spatial patterns, such as differences between coastal and inland regions for example.
The descriptions of Tables 1 and 2 are very extensive. Consider focusing on results that are significant from a broader perspective. Are the SS spatially correlated? Where do they occur, and do they exhibit clear spatial patterns?
3.2 Characteristics of CWPDs intensity
This section largely repeats the previous one.
It is expected to have differences in the intensities as the selection of the extreme events is based on different thresholds.
I suggest focusing on spatial patterns and on comparing similarities and differences between the frequency and intensity of CWPDs.
3.3 Dependence of concurrent days on individual ones
497–503: This information is already evident from the figure and could be removed.
The analysis itself is interesting, but the description and interpretation of the results need improvement. For example, what drives these compound events? What wind and precipitation intensities are typically observed? Are there regional differences or identifiable spatial patterns?
4 Discussion
The discussion relies heavily on previous studies and sometimes feels disconnected from the results presented here. A stronger interpretation of the study’s own findings, and a clearer link to existing literature, is needed.
589–593: These results are expected given the use of percentile-based thresholds (see previous comments).
607–630: This section discusses seasonal differences based on previous studies, but the results presented here do not explicitly show seasonal analyses. Therefore, it is unclear how conclusions about tropical cyclones in summer (line 625) are supported. Including seasonal results (e.g., summer vs. winter and threshold dependence) would strengthen the discussion.
632–643: This paragraph discusses regional differences across China, but some of these regions were not introduced earlier. Including them in Figure 1 or an additional supplementary figure would be beneficial. Furthermore, how do these regional characteristics relate to your results? This type of analysis would be more informative than extensively describing figures in the Results section.
682: “winter ground wind speed” - Seasonal differences are not clearly reflected in the results
Several discussion paragraphs are disconnected (e.g., lines 685–691 and 716–718), and some sections repeat results (e.g., lines 725–730).
706: “this is consistent with our findings” - Please specify which results or figures support this statement.
719–722: What is the spatial relationship or pattern being referred to here?
5 Conclusions
The conclusions would benefit from a clearer summary of the identified trends. In a concise manner, please specify how the frequency and intensity of CWPDs are changing. For example, can we expect more or fewer events, and will they be more or less intense? What are the main drivers? Are there specific regions with distinct patterns or trends?
Minor comments
76: “precipitation extremes as the typical” - grammar error; likely “are” instead of “as”.
99: “in coastal regions those frequently” - unclear use of “those”.
102: “WRF” - abbreviation not previously defined.
104: “showed that the areas that the most frequent” - should be “that are the most frequent”.
111–114: Sentence is unclear and should be rephrased.
149 / Figure 1: What does “the same as below” mean?
177: Sentence is unclear; likely missing verbs (e.g., “is defined when…”).
200 & 202: Please clarify what is meant by “sequences”.
493: “with a speed of 0.33 days per decade” – use “rate” instead of “speed”
494: “non-significantly at a speed of 2.01 days per decade” – use “rate” instead of “speed”
Figure 8: Title, description, and legend could be improved.
523: “The scopes of the ratio of CWPDs to extreme” – what do you mean with “scopes”? please clarify.
Figure 9: Improve title, description, and legend; explain the meaning of “+” and “−”.
622: “ARs-related and cyclone-ARs-related events during 1979 to 2020, CWPDs in SC showed decreasing trends for frequency and intensity in winter related to the decreasing of ARs and cyclone-ARs events” - reference needed
628: use “tropical cyclone” instead of “TC”, abbreviations are heavily used in the manuscript, try to avoid them when it is not needed.
676: “PDO” - abbreviation not defined.
736: “causing huge losses in multiple aspects including water security, food security and human health.” - reference needed.

Citation: https://doi.org/10.5194/egusphere-2025-5893-RC2
- AC2: 'Reply on RC2', K.M. Wen, 06 Mar 2026
  
  This manuscript investigates compound wind and precipitation events (CWPDs) using an extensive network of meteorological stations across China. Both the frequency and intensity of compound events are evaluated using different percentile-based threshold combinations. Overall, the manuscript is highly descriptive but lacks a more comprehensive interpretation of the results.
  
  The connection between the methodology and the results should be strengthened. In particular, the selection of the time window and percentile thresholds applied to the time series requires clearer justification, as does the rationale for combining specific wind and precipitation thresholds. The manuscript would also benefit from additional analyses addressing the seasonality of CWPDs, as well as spatial correlations and/or trends.
  Thank you very much for the comments. We have made a new revision of the manuscript, referring to the comments and suggestion.
  
  Comments
  
  1 Introduction
  
  (1) There is a strong emphasis on compound wind–precipitation events in coastal areas throughout the introduction; however, the stated objective clearly indicates that the focus is mainland China. This inconsistency should be addressed.
  Thank you for the suggestion. In the introduction section of this revised manuscript, we have added the contents of compound wind–precipitation events in mainland China: “ ”.
  
  (2) 119: “so the results are less accurate than ground-based observations” - This statement is not necessarily correct. Reanalysis and model data are typically calibrated and validated against ground-based observations.
  Thank you for the suggestion. In the introduction section of the revised manuscript, we have redescribed the flaws of reanalysis products: In some regions with complex terrain (e.g., mountainous areas), reanalysis datas are not suitable as a benchmark for studying compound wind speed and precipitation extremes (Zscheischler et al., 2020b; Zhang et al., 2021). Observational records is crucial for accurately, comprehensively investigating the authentic characteristics of compound wind speed and precipitation extremes”.
  
  2 Data
  
  (3) This section should be titled “Data and Methods” rather than only “Data.”
  Thank you for the suggestion. In the revised manuscript, we have changed the section title of “Data” into “Data and Methods”.
  
  (4) In several cases, the methods lack sufficient explanation or justification for key choices (e.g., threshold selection). Additionally, some methods are mentioned without even a brief explanation, while others are described in detail. The linkage between the methods and the results could be improved.
  Thank you for the suggestion. In the revised manuscript, we have made a new revision of the data and methods section, referring to the comments and suggestion.
  
  2.1 Data
  
  (5) 144: “The missing records of wind speed and precipitation in any station were retained.” - How were these missing records handled in the analysis?
  Thank you for the comment. In the data and methods section of this revised manuscript, we have redescribed the data processing procedure: “The relevant data should coexist at a given station in the same time series to facilitate the study of compound wind speed and precipitation extreme characteristics. The maximum wind speed records at many stations had missing values before 1981, so to ensure the consistency and integrity of the daily sequences, if the missing rate of daily records at a station exceeded 20% of the total daily records from 1981 to 2022, the meteorological station would be removed from the research dataset. Ultimately, 1686 meteorological stations were selected for analysis (Fig. 1). Among the 1686 stations, there were 1292 stations with the missing rate of daily maximum wind speed records of less than 3%, and 1654 stations with that of daily precipitation records of less than 3% during 1981–2022. The missing values of wind speed and precipitation in any station were replaced by the records from the nearest station on the same day”.
  
  (6) Overall, a clearer overview of the dataset is needed, including the number of days with wind events, precipitation events, and CWPDs.
  Thanks for the suggestion. In the data and methods section of this revised manuscript, we have redescribed the dataset: “Two long-term daily historical observation datasets of daily maximum wind speed and daily total precipitation from 2481 meteorological stations over mainland China were used. These datasets were from the National Meteorological Information Center of the China Meteorological Administration (NMIC/CMA) (http://data.cma.cn/). The data had undergone strict quality control and homogeneity testing before release and had been widely applied in related research (Li et al., 2015; Yu and Zhai, 2020). “Daily maximum wind speed” refers to the maximum value of the 10-min average wind speed (10 m height) in a given day.
  The relevant data should coexist at a given station in the same time series to facilitate the study of compound wind speed and precipitation extreme characteristics. The maximum wind speed records at many stations had missing values before 1981, so to ensure the consistency and integrity of the daily sequences, if the missing rate of daily records at a station exceeded 20% of the total daily records from 1981 to 2022, the meteorological station would be removed from the research dataset. Ultimately, 1686 meteorological stations were selected for analysis (Fig. 1). Among the 1686 stations, there were 1292 stations with the missing rate of daily maximum wind speed records of less than 3%, and 1654 stations with that of daily precipitation records of less than 3% during 1981–2022. The missing values of wind speed and precipitation in any station were replaced by the records from the nearest station on the same day”.
  
  2.2 Definition of compound extremes
  
  (7) 166: The justification of the 15-day time window is missing
  Thanks for the comments. We used 15-day moving window to establish the thresholds series reference the methods of Wang et al. (2021) and Luo et al. (2022). The adoption of a 15-day sliding time window can eliminate the systematic errors in the data and other unreasonable human factors, thus making the calculation of thresholds more objective. Otherwise, the results based on 15-day sliding time window has been verified to be accurate (Yang et al., 2025).
  
  References:
  Wang, J., Y. Chen, W. Liao, et al., 2021: Anthropogenic emissions and urbanization increase risk of compound hot extremes in cities. Nat. Clim. Chang., 11, 1084–1089, https://doi.org/10.1038/s41558-021-01196-2.
  Luo, M., N. C. Lau, and Z. Liu, 2022: Different mechanisms for daytime, nighttime, and compound heatwaves in southern China. Weather and Climate Extremes, 36, 100449, https://doi.org/10.1016/j.wace.2022.100449.
  Yang, Y., and C. Yuan, 2025: Causality of compound extreme heat-precipitation events in Northeastern China. Atmospheric Research, 107975, https://doi.org/10.1016/j.atmosres.2025.107975.
  
  (8) 168-169: This sentence is not clear
  Thanks for the suggestion. In the revised manuscript, because this sentence is unnecessary, we have deleted this sentence.
  
  (9) 173: The choice of percentile thresholds (85th, 90th, 95th, 98th) requires justification
  Thanks for the suggestion. In this revised manuscript, we have referred to more references to illustrate their suitability in capturing compound extremes and the percentiles chosen are designed to serve as a trade-off between extremeness and sample size.
  
  (10) 179: The rationale for the selected threshold combinations is missing.
  Thanks for the suggestion. In this revised manuscript, we have added the explanations for selected the thresholds: “According to existing studies (Li et al., 2022; Zhang et al., 2021; Martius et al., 2016; Yaddanapudi et al., 2022; Zhou and Liu, 2018), four combinations of wind speed and precipitation thresholds were used in this study, including 85th wind speed and 85th precipitation (W85∩P85), 90th wind speed and 90th precipitation (W90∩P90), 95th wind speed and 95th precipitation (W95∩P95), 98th wind speed and 98th precipitation (W98∩P98). The percentiles chosen here were designed to serve as a trade-off between extremeness and sample size (Ridder et al., 2020; Lai et al., 2021; Hao et al., 2018) and were used to compare the effects of different thresholds on study results”.
  
  (11) 196: “ultimately isolating the top 10 concurrent extremes for further analysis.” – Is this actually done? The results appear to include all events. If not, a clearer explanation is needed.
  Thanks for the suggestion. For the the top 10 daily CWPDI didn’t analyze in present study, in the data and methods section of this revised manuscript, we have deleted the sentence: “We summed the CWPDI values of all stations in China on each day (1981–2022) to determine the daily cumulative CWPDI values in descending order, ultimately isolating the top 10 concurrent extremes for further analysis”.
  
  2.3 Regional division and linear trend
  
  (12) 199: It would be helpful to indicate how many stations are included in each region (e.g., in Figure 2).
  Thanks for the suggestion. In the 2.3 section of the revised manuscript, we have added the explanations of how many stations are included in each area.
  
  (13) 209: Briefly describe the Sen–Theil method.
  Thanks for the suggestion. In the 2.3 section of the revised manuscript, we have briefly described the Sen–Theil method: “Sen-Theil method assumes a linear trend and represents the quantification of temporal change. Because of its robustness against the effect of outliers, this test is preferred over linear regression in hydro-meteorological investigations (Zhang et al., 2005)”.
  
  Reference:
  Zhang, X., Hegerl, G., Zwiers, F. W., & Kenyon, J. (2005). Avoiding Inhomogeneity in Percentile-Based Indices of Temperature Extremes. Journal of Climate, 18, 1641-1651. https://doi.org/10.1175/JCLI3366.1
  
  (14) 219: Why was the W85∩P85 threshold combination selected?
  Thank you for the comment. In the 2.3 section of the revised manuscript, we have explained the reasons for choosing the threshold of W85∩P85: “For a nationwide analysis, a higher threshold will result in a smaller sample size in inland areas of the country, and the threshold of W85∩P85 can ensure an adequate sample size to reflect the differences in CWPD variations across different regions. Therefore, the threshold of W85∩P85 was used for further explore seasonal characteristics of CWPDs, and the relationship between extreme wind speed, extreme precipitation and CWPDs”.
  
  (15) 220: Please explain how the frequency ratios are defined and calculated.
  Thanks for the suggestion. In the 2.3 section of the revised manuscript, we have explained how the the frequency ratios were calculated: “The frequency ratios were calculated through CWPDs divided by the corresponding precipitation and wind speed extremes to evaluate the linkage between individual and concurrent extremes (Zhang et al., 2021)”.
  
  3 Results
  
  (16) Hainan Province is mentioned multiple times without being introduced. It should be clearly identified in Figure 1.
  Thanks for the suggestion. In the revised manuscript, we have clearly marked the location of Hainan Province in Figure 1.
  
  (17) Information regarding threshold intensity is missing. For example, which actual wind speeds and precipitation amounts correspond to the different percentiles? Are there clear spatial patterns in these intensities?
  Thanks for the suggestion. In the results section the revised manuscript, we have added the content of threshold intensity: “The distribution of different percentile wind speed thresholds is generally consistent, all showing the characteristic that the thresholds in the northern China and SWC are larger, while those in the YR, SC, and eastern and southern parts of NC are smaller (Fig. S1). In most of YR and SC, the 85th percentile wind speeds thresholds were 3.5-6 m/s, the 90th percentile thresholds were 3.9-6 m/s, the 95th percentile thresholds were 4.3-7 m/s, and the 98th percentile thresholds were 4.8-8 m/s. In northeastern YR and mid-southern NC, the 85th and 90th percentile wind speeds thresholds were 6-8m/s, the 95th percentile thresholds were 7-9 m/s, and the 98th percentile thresholds were 8-11 m/s. In NEC, parts of mid-eastern and northwestern NWC, parts of mid-eastern SWC, the 85th and 90th percentile wind speeds thresholds were 8-10m/s, the 95th percentile thresholds were 9-12 m/s, and the 98th percentile thresholds were 11-14 m/s. In northern NC, northern SWC, southwestern NWC, the 85th percentile wind speeds thresholds were 10-18m/s, the 90th percentile thresholds were 10-19.4 m/s, the 95th percentile thresholds were 12-22 m/s, and the 98th percentile thresholds were 14-24.6 m/s (Fig. S1a, S1c, S1e, S1g).
  The distribution of different percentile precipitation thresholds was quite similar, all showing a decreasing trend from the southeast to the northwest. However, the maximum value area of the 98th percentile precipitation threshold is mainly concentrated in the coastal areas of South China, while the maximum value areas of the other percentile thresholds not only cover the coastal areas of SC but also include some parts of the mid-eastern YR (Fig. S1). In northern NC, NWC, central and northern SWC, northwestern NEC, the 85th percentile precipitation thresholds were 1.4-7 mm, the 90th percentile thresholds were 1.7-8 mm, the 95th percentile thresholds were 2.3-10 mm, and the 98th percentile thresholds were 3-15 mm. In central and southwestern NC, most NEC, parts of eastern SWC, parts of northwestern NWC, the 85th percentile precipitation thresholds were 7-11 mm, the 90th percentile thresholds were 8-14 mm, the 95th percentile thresholds were 10-20 mm, and the 98th percentile thresholds were 15-30 mm. In southwestern NEC, parts of eastern and southern NC, western and northwestern YR, southeastern SWC, the 85th percentile precipitation thresholds were 11-15 mm, the 90th percentile thresholds were 14-21 mm, the 95th percentile thresholds were 20-30 mm, and the 98th percentile thresholds were 30-45 mm. In mid-eastern YR, parts of southeastern NC, parts of western and northern SC, the 85th percentile precipitation thresholds were 15-20 mm, the 90th percentile thresholds were 21-28 mm, the 95th percentile thresholds were 30-40 mm, and the 98th percentile thresholds were 45-60 mm. In parts of mid-eastern YR, parts of southern and southeastern SC, the 85th percentile precipitation thresholds were 20-27.8 mm, the 90th percentile thresholds were 28-38.2 mm, the 95th percentile thresholds were 40-61 mm, and in parts of southern and southeastern SC, the 98th percentile thresholds were 60-92.3 mm (Fig. S1b, S1d, S1f, S1h)”.
  
  (18) Figures and tables are described extensively, but there is limited discussion of the physical mechanisms or spatial patterns underlying the results.
  Thanks for the suggestion. In the results section the revised manuscript, we have restructure the results contents, added the spatial patterns and physical mechanisms descriptions.
  
  (19) An analysis by region may be more informative.
  Thanks for the suggestion. In the results section the revised manuscript, we have redescribed the results contents from the perspective of regions.
  
  3.1 Characteristics of CWPDs frequency
  
  (20) Figure 2: Use consistent color scales for all four threshold combinations. The region names are difficult to read; consider moving them to the foreground or outside the map.
  Thank you for the comment. In the revised manuscript, we have moved the region names to outside the map. Due to the significant differences in frequency and intensity of different thresholds across the country, in order to reflect spatial heterogeneity, different color scales have been adopted for different thresholds.
  
  (21) It is expected that similar spatial patterns will emerge across different thresholds, as they are percentile-based. Much of the text reiterates what is already visible in the figures and becomes repetitive when discussing multiple threshold combinations.
  Thanks for the suggestion. In the revised manuscript, we have condensed the description of the results of four thresholds and added the content of spatial patterns underlying different thresholds.
  
  (22) Instead, the discussion should focus more on spatial patterns, such as differences between coastal and inland regions for example.
  Thanks for the suggestion. In the results section of the revised manuscript, we have analyzed the results from the perspective of spatial patterns, and added the comparative analysis between the coastal areas and the inland regions.
  
  (23) The descriptions of Tables 1 and 2 are very extensive. Consider focusing on results that are significant from a broader perspective. Are the SS spatially correlated? Where do they occur, and do they exhibit clear spatial patterns?
  Thanks for the suggestion. In the 3.1 section of the revised manuscript, we have described Tables 1 and 2 combined with the spatial distribution, and analyzed from spatial patterns perspective.
  
  3.2 Characteristics of CWPDs intensity
  
  (24) This section largely repeats the previous one.
  Thanks for the suggestion. In the revised manuscript, we have reduced the repetitive contents of the 3.2 section, described the results more in terms of spatial patterns, and added the contrast with the frequency part.
  
  (25) It is expected to have differences in the intensities as the selection of the extreme events is based on different thresholds.
  Thanks for the suggestion. In the revised manuscript, we have reduced the redundant contents, and highlighted the key points and described the spatial distribution patterns by combining with the table.
  
  (26) I suggest focusing on spatial patterns and on comparing similarities and differences between the frequency and intensity of CWPDs.
  Thanks for the suggestion. In the 3.2 section of the revised manuscript, we have described the results focusing on spatial patterns, and compared the similarities and differences between the frequency and intensity of CWPDs: “For mainland China as a whole, from 1981 to 2022, the frequencies of the four thresholds all showed a downward trend. As the threshold increases, the downward trend becomes smaller. The intensities of the W85∩P85 and W90∩P90 thresholds showed a significant downward trend, while the intensities of the W95∩P95 and W98∩P98 thresholds did not change significantly. When looking at the time periods separately, both frequency and intensity showed a weakening trend from 1981 to 2010, and an increasing trend from 2011 to 2022. In terms of spatial distribution, the trends of frequency and intensity are generally consistent, and both are the area showing a decreasing trend larger than the area showing an increasing trend. As the threshold increases, the spatial trend distribution of CWPDs frequency does not change significantly at W85∩P85, W90∩P90, and W95∩P95 thresholds. When the threshold was raised to W98∩P98 percentiles, the area of frequency showing an increasing trend significantly expands. The spatial trend distribution of CWPDs intensity does not change much at W85∩P85 and W90∩P90 thresholds, when the thresholds were raised to W95∩P95 and W98∩P98 percentiles, the area of intensity showing an enhanced trend significantly expands. In terms of the W98∩P98 threshold, the stations with an increasing trend in intensity and those with a decreasing trend are almost the same”.
  
  3.3 Dependence of concurrent days on individual ones
  
  (27) 497–503: This information is already evident from the figure and could be removed.
  Thanks for the suggestion. In the 3.3 section of the revised manuscript, we have deleted this information.
  
  (28) The analysis itself is interesting, but the description and interpretation of the results need improvement. For example, what drives these compound events? What wind and precipitation intensities are typically observed? Are there regional differences or identifiable spatial patterns?
  Thanks for the suggestion. In the 3.3 section of the revised manuscript, we have added the descriptions of spatial patterns and the discussions of physical mechanisms.
  
  4 Discussion
  
  (29) The discussion relies heavily on previous studies and sometimes feels disconnected from the results presented here. A stronger interpretation of the study’s own findings, and a clearer link to existing literature, is needed.
  Thank you for the comment. In the discussion section of the revised manuscript, we have added the interpretations of the findings, and clearly distinguish the findings from the existing literature.
  
  (30) 589–593: These results are expected given the use of percentile-based thresholds (see previous comments).
  Thanks for the suggestion. In the discussion section of the revised manuscript, we have deleted this informations, and hightlights the meaningful findings.
  
  (31) 607–630: This section discusses seasonal differences based on previous studies, but the results presented here do not explicitly show seasonal analyses. Therefore, it is unclear how conclusions about tropical cyclones in summer (line 625) are supported. Including seasonal results (e.g., summer vs. winter and threshold dependence) would strengthen the discussion.
  Thanks for the suggestion. In the results section of the revised manuscript, we have added the seasonal results (Figs. 5, 9, S3 and S5).
  
  (32) 632–643: This paragraph discusses regional differences across China, but some of these regions were not introduced earlier. Including them in Figure 1 or an additional supplementary figure would be beneficial. Furthermore, how do these regional characteristics relate to your results? This type of analysis would be more informative than extensively describing figures in the Results section.
  Thanks for the suggestion. In the revised manuscript, we have added indications for special areas in figure 1, and compared these regional characteristics to the results.
  
  (33) 682: “winter ground wind speed” - Seasonal differences are not clearly reflected in the results
  Thanks for the suggestion. In the results section of the revised manuscript, we have added the seasonal results (Figs. 5, 9, S3 and S5).
  
  (34) Several discussion paragraphs are disconnected (e.g., lines 685–691 and 716–718), and some sections repeat results (e.g., lines 725–730).
  Thanks for the suggestion. In the revised manuscript, we have restructure the discussion section into: (a) key findings, (b) interpretation linked to results and (c) implications and limitations, and deleted the repeat results.
  
  (35) 706: “this is consistent with our findings” - Please specify which results or figures support this statement.
  Thanks for the suggestion. In the revised manuscript, we have specified the results that support this statement, and placed the content of comparison with literature to the results section.
  
  (36) 719–722: What is the spatial relationship or pattern being referred to here?
  Thanks for the suggestion. In the revised manuscript, we have clearly stated the spatial pattern: “Different percentile thresholds had effects on the spatial pattern and change trend of CWPDs. The frequencies of CWPDs identified by the W90∩P90 (Fig. 2b) were comparable to those identified by the W85∩P85 (Fig. 2a), but the former resulted in smaller frequencies than the latter in almost all regions. In general, southeastern SC, Hainan Province, northwestern YR, some scattered areas in the central and eastern YR were areas where the CWPDs were frequent, and CWPDs were less in most of NWC, SWC, NEC and NC (Fig. 2). The CWPDs frequencies under the four thresholds all showed decreasing trends in most areas of mainland China during 1981 to 2022, as the threshold increases, the trends of decreased for CWPDs frequencies decreased obviously, the spatial trend distribution of CWPDs frequency does not change significantly at W85∩P85, W90∩P90, and W95∩P95 thresholds. When the threshold was raised to W98∩P98 percentiles, the area of frequency showing an increasing trend significantly expands (Fig. 4). With the increase of the threshold, the area of CWPDs with stronger intensity further reduced. The CWPDs intensities were more severe in southeastern coastal areas, including eastern coastal areas of YR, mid-eastern SC, parts of eastern SWC, parts of central NEC, parts of northwestern NWC and mid-northern in Hainan Province (Fig. 6). From 1981 to 2022, the CWPDs intensities under the four thresholds all showed decreasing trends in most areas of mainland China except for parts of central SC, a few scattered areas of YR, several scattered areas of NC and NEC, a few scattered areas of NWC, individual areas of eastern SWC. As the thresholds increased, the trends of weakened for CWPDs intensities decreased and the areas with a slight strengthening trend expanded (Fig. 8). The spatial trend distribution of CWPDs intensity does not change much at W85∩P85 and W90∩P90 thresholds, when the thresholds were raised to W95∩P95 and W98∩P98 percentiles, the area of intensity showing an enhanced trend significantly expands. In terms of the W98∩P98 threshold, the stations with an increasing trend in intensity and those with a decreasing trend are almost the same”.
  
  5 Conclusions
  
  (37) The conclusions would benefit from a clearer summary of the identified trends. In a concise manner, please specify how the frequency and intensity of CWPDs are changing. For example, can we expect more or fewer events, and will they be more or less intense? What are the main drivers? Are there specific regions with distinct patterns or trends?
  Thanks for the suggestion. In the revised manuscript, we have rewrited the conclusions and clearly summarize the discoveries: “We investigated the observed CWPDs patterns in the past 42 years over mainland China by using the thresholds that vary with dates and stations, which could accurately provide information of CWPDs on different time scales and periods for different locations under the background of global warming. Our analysis identified the hotspot locations of CWPDs were in northwestern and eastern coastal areas of YR, mid-eastern and southeastern SC, Hainan Province, parts of eastern SWC, parts of central NEC, parts of northwestern NWC, which complements previous studies (Bevacqua et al., 2020; Couasnon et al., 2020; Martius et al., 2016). Both the frequencies and the intensities of CWPDs changed obvious around early-to-mid 2010s under four different thresholds, showed a downward trend in 1981–2010 followed by an upward trend in 2011–2022, indicating that high-intensity CWPDs will occur frequently in the future, especially in the hotspot locations, this is of great practical significance in guiding these regions to formulate corresponding climate adaptation and mitigation strategies. The CWPDs frequencies decreased during 1981-2010 were mainly caused by the decreased of extreme wind speed days, but the CWPDs frequencies increased during 2011-2022 were caused by the increased of both extreme wind speed days and extreme precipitation days. The changes of extreme wind speed days were consistent with those of CWPDs during 1981–2010 and 2011–2022, but the change of extreme precipitation days lacked correspondence with that of CWPDs in the two periods, especially in the previous period.
  Different percentile thresholds had effects on the spatial pattern and change trend of CWPDs. As the threshold increases, the CWPDs frequencies under the four thresholds all showed decreasing trends in most areas of mainland China during 1981 to 2022, the trends of decreased for CWPDs frequencies decreased obviously, the spatial trend distribution of CWPDs frequency does not change significantly at W85∩P85, W90∩P90, and W95∩P95 thresholds. When the threshold was raised to W98∩P98 percentiles, the area of frequency showing an increasing trend significantly expands. The spatial trend distribution of CWPDs intensity does not change much at W85∩P85 and W90∩P90 thresholds, when the thresholds were raised to W95∩P95 and W98∩P98 percentiles, the area of intensity showing an enhanced trend significantly expands. In terms of the W98∩P98 threshold, the stations with an increasing trend in intensity and those with a decreasing trend are almost the same.
  Our findings provide valuable information (e.g., hotspot locations, intensity analysis, and sensitivity analysis) related to the observed CWPDs patterns over mainland China. The findings will enhancement future researches on CWPDs dominated by different weather systems, such as tropical cyclones in southeastern coastal areas of mainland China (Rodgers et al., 1994) and stronger extratropical cyclones in Jianghuai region and Northeast region of mainland China (Raveh-Rubin & Wernli, 2015). The findings from our study can help reduce the risk from concurrent events for various stakeholders, such as coastal communities, utility industries, agriculture agency, emergency management agency, and the insurance sector”.
  
  Minor comments
  
  (38) 76: “precipitation extremes as the typical” - grammar error; likely “are” instead of “as”.
  Thanks for the suggestion. In the revised manuscript, we have replaced "as" with "are".
  
  (39) 99: “in coastal regions those frequently” - unclear use of “those”.
  Martius et al. (2016) made the first global quantification of compound precipitation and wind extremes and found that compound extremes mainly happened in the coastal areas where affected by tropical cyclones (TCs) frequently.
  Thanks for the suggestion. In the revised manuscript, we have deleted “those” and revised "Martius et al. (2016) made the first global quantification of compound precipitation and wind extremes, indicating that compound precipitation and wind extremes mainly happen in coastal regions those frequently affected by tropical cyclones (TCs)" with "Martius et al. (2016) made the first global quantification of compound precipitation and wind extremes, indicating that compound precipitation and wind extremes mainly happen in coastal regions frequently affected by tropical cyclones (TCs)".
  
  (40) 102: “WRF” - abbreviation not previously defined.
  Thanks for the suggestion. In the revised manuscript, we have explained WRF into “Weather Research and Forecasting Model”.
  
  (41) 104: “showed that the areas that the most frequent” - should be “that are the most frequent”.
  Thanks for the suggestion. In the revised manuscript, we have replaced “the areas that the most frequent” with “the areas that are the most frequent”.
  
  (42) 111–114: Sentence is unclear and should be rephrased.
  Thanks for the suggestion. In the revised manuscript, we have replaced “Waliser and Guan (2017) found strong connection between individual extreme wind or extreme precipitation and atmospheric rivers (ARs) across the vast midlatitudes zone, indicated that of ARs as a potential driver of compound wind and precipitation extremes” with “Waliser and Guan (2017) discerned a strong relation between individual extreme wind or precipitation and atmospheric rivers (ARs) across broad swathes of midlatitudes from a univariate perspective, and indicated that ARs were a potential driver of compound wind and precipitation extremes”.
  
  (43) 149 / Figure 1: What does “the same as below” mean?
  Thanks for the suggestion. In the revised manuscript, we have deleted the “the same as below”.
  
  (44) 177: Sentence is unclear; likely missing verbs (e.g., “is defined when…”).
  Thanks for the suggestion. In the revised manuscript, we have added the verb “was” to the sentence: “One compound day was defined when the daily maximum wind speed higher than the percentile and the daily precipitation amount higher than the percentile at the same time”.
  
  (45) 200 & 202: Please clarify what is meant by “sequences”.
  Thanks for the suggestion. In the revised manuscript, we have replaced “regional sequences” with “regional average annual extremes series”.
  
  (46) 493: “with a speed of 0.33 days per decade” – use “rate” instead of “speed”
  Thanks for the suggestion. In the revised manuscript, we have replaced “speed” with “rate”.
  
  (47) 494: “non-significantly at a speed of 2.01 days per decade” – use “rate” instead of “speed”
  Thanks for the suggestion. In the revised manuscript, we have replaced “speed” with “rate”.
  
  (48) Figure 8: Title, description, and legend could be improved.
  Thanks for the suggestion. In the revised manuscript, we have improved the title, description, and legend of figure 8.
  
  (49) 523: “The scopes of the ratio of CWPDs to extreme” – what do you mean with “scopes”? please clarify.
  Thanks for the suggestion. In the revised manuscript, we have replaced “scopes” with “areas”, and rewrited the sentence: “The areas where the ratios of CWPDs to precipitation extremes were showing an increasing trend has expanded compared to the period of 1981-2010 in most regions”.
  
  (50) Figure 9: Improve title, description, and legend; explain the meaning of “+” and “−”.
  Thanks for the suggestion. In the revised manuscript, we have improved the title, description, and legend of figure 9, and explain the meaning of “+”: “Positive sign denotes the trend significant at 95% confidence leve”.
  
  (51) 622: “ARs-related and cyclone-ARs-related events during 1979 to 2020, CWPDs in SC showed decreasing trends for frequency and intensity in winter related to the decreasing of ARs and cyclone-ARs events” - reference needed
  Thanks for the suggestion. In the revised manuscript, we have added reference for this content.
  
  (52) 628: use “tropical cyclone” instead of “TC”, abbreviations are heavily used in the manuscript, try to avoid them when it is not needed.
  Thanks a lot. In the revised manuscript, we have use “tropical cyclone” instead of “TC”, and other abbreviations, except for the regional ones, have been replaced with their full names.
  
  (53) 676: “PDO” - abbreviation not defined.
  Thanks a lot. In the revised manuscript, we have use “Pacific Decadal Oscillation” instead of “PDO”.
  
  (54) 736: “causing huge losses in multiple aspects including water security, food security and human health.” - reference needed.
  Thanks for the suggestion. In the revised manuscript, we have added reference for this content.
  
  Thanks again for your comments and suggestions.
  
  Citation: https://doi.org/10.5194/egusphere-2025-5893-AC2
- AC3: 'Reply on RC2', K.M. Wen, 06 Mar 2026
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-5893/egusphere-2025-5893-AC3-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2025-5893-AC3

Kangmin Wen and Jun Shi

Supplement

https://doi.org/10.5194/egusphere-2025-5893-supplement

Kangmin Wen and Jun Shi

Viewed

Total article views: 401 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
295	72	34	401	52	48	29

HTML: 295
PDF: 72
XML: 34
Total: 401
Supplement: 52
BibTeX: 48
EndNote: 29

Views and downloads (calculated since 04 Dec 2025)

Month	HTML	PDF	XML	Total
Dec 2025	109	37	18	164
Jan 2026	68	11	3	82
Feb 2026	39	15	6	60
Mar 2026	79	9	7	95

Cumulative views and downloads (calculated since 04 Dec 2025)

Month	HTML	PDF	XML	Total
Dec 2025	109	37	18	164
Jan 2026	68	11	3	82
Feb 2026	39	15	6	60
Mar 2026	79	9	7	95

Viewed (geographical distribution)

Total article views: 367 (including HTML, PDF, and XML) Thereof 367 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 25 Mar 2026

Short summary

Based on the daily maximum wind speed and precipitation data, the spatial patterns and inter-annual variations of concurrent wind speed and precipitation days under different thresholds, their spatial trends, regional differences in different periods (1981–2010 and 2011–2022), as well as the spatio-temporal characteristics and changes of the influence or contribution of wind speed days and precipitation days to the occurrence of concurrent wind speed and precipitation days were analyzed.


Total:	0
HTML:	0
PDF:	0
XML:	0